Front Cover

Group Title: Circular
Title: Agricultural water quality sampling strategies
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00014459/00001
 Material Information
Title: Agricultural water quality sampling strategies
Series Title: Circular
Physical Description: 10 p. : ill. ; 28 cm.
Language: English
Creator: Izuno, Forrest T
Bottcher, A. B
Davis, W.A ( Winston A )
Publisher: Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, University of Florida
Place of Publication: Gainesville Fla
Publication Date: 1992
Subject: Agricultural pollution -- Florida   ( lcsh )
Water quality -- Measurement   ( lcsh )
Water -- Pollution -- Measurement   ( lcsh )
Genre: government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
non-fiction   ( marcgt )
Bibliography: Includes bibliographical references (p. 10).
Statement of Responsibility: F.T. Izuno, A.B. Bottcher, and W.A. Davis.
General Note: Title from cover.
General Note: "January 1992."
 Record Information
Bibliographic ID: UF00014459
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: ltqf - AAA6898
ltuf - AJG5668
oclc - 26846438
alephbibnum - 001752711

Table of Contents
    Front Cover
        Front Cover 1
        Front Cover 2
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
Full Text

I 1O
January 1992


Circular 1036

arl e6

Agricultural Water Qua]

Sampling Strategies

F.T. Izuno, A.B. Bottcher, and W.A. Davis


Florida Cooperative Extension Service
Institute of Food and Agricultural Sciences
University of Florida
John T. Woeste, Dean

F.T. Izuno, Associate Professor, Everglades Research & Education Center, University of Florida, IFAS; A.B. Bottcher, Professor, Ag-
ricultural Engineering Department, University of Florida, IFAS; W.A. Davis, Sr. Chemist, Everglades Research & Education Center,
University of Florida, IFAS.

Agricultural nutrients, primarily nitrogen (N)
and phosphorus (P), in drainage water from farms
in the Everglades Agricultural Area (EAA) are al-
leged to be contributing to the eutrophication of
Lake Okeechobee and undesirable changes in the
Water Conservation Areas (WCAs) and the Ever-
glades National Park (ENP) ecosystems (Florida
Department of Administration, 1976; LOTAC I,
1986; LOTAC II, 1990; SFWMD, 1990). The ques-
tion of whose responsibility it is to alleviate the flux
of nutrients to those protected environments is
presently (1991) being debated. There are many
indications that much of the physical and financial
burden will be placed on the agricultural commu-
nity (SFWMD, 1990).

Nitrogen and P concentrations in EAA farm
drainage water are higher than background levels
in each of the three ecosystems (Lake Okeechobee,
WCAs, ENP), and are therefore considered to be
environmental threats. Phosphorus has been iden-
tified as the limiting nutrient (SFWMD, 1990). In
other words, increasing P concentrations will accel-
erate eutrophication of area waters and further
stimulate the proliferation of undesirable plant
growth, while decreasing P concentrations will
have the opposite effects.

The high nutrient levels in EAA drainage can be
partially attributed to the high water table organic
soils of the EAA which are shallow, drained, and
subject to biological oxidation and mineralization of
both N and P. Fertilizer P applications vary
widely, from zero to several hundred lbs/ac/yr, de-
pending on the crop and P available in the soils.
Therefore, fertilization practices have a variable
impact on P concentrations throughout the EAA.
Nitrogen fertilizer is rarely used. Rainfall events
in south Florida are characteristically of high vol-
ume and intensity, necessitating drainage to off-
farm canals. High nutrient concentrations and
large drainage volumes translate to high nutrient
loading rates to drainage water receiving areas.

Attempts to reduce area nutrient loadings
should logically begin at the farm level. The Sur-
face Water Improvement and Management (SWIM)
Plan (SFWMD, 1990) suggests implementing on-
farm Best Management Practices (BMPs) to reduce
on-farm P concentrations and loadings. After low-
ering P concentrations and loadings at the farm
level, regional Stormwater Treatment Areas (STAs)
would be used to further polish the agricultural
drainage water. To assure the proper management

and optimum effectiveness of BMPs, water quality
monitoring will be a necessary activity for the
growers in the EAA.

Large variations in P concentrations occur be-
tween farms, seasonally, and within individual
drainage events on a farm (Izuno et al., 1990a;
Izuno et al. 1991b; and Izuno and Bottcher; 1991).
Therefore, simply collecting and analyzing a water
sample occasionally will provide little useful infor-
mation. A meaningful farm water quality monitor-
ing program requires that a comprehensive moni-
toring program will need to be developed and
implemented at each farm location. Because the
physical characteristics, water management, and
cultural practices differ among farms, each site
could potentially require a unique program. The
purpose of this publication is to provide general in-
formation to consider when developing a water
quality monitoring program for an individual farm.

Nutrient parameters to monitor
A water sample can be analyzed for numerous
chemical properties. This publication will be lim-
ited to a discussion of N and P, with the emphasis
being on P. Total-N (TN) and Total-P (TP) mea-
surements account for all N and P, respectively, in
organic and inorganic forms, measurable using
standard chemical analysis procedures. Total-N
cannot be measured directly, and is therefore deter-
mined as the sum of Total Kjeldahl Nitrogen (TKN)
(Izuno et al., 1991a) and Nitrate (NO3). Total Dis-
solved Kjeldahl Nitrogen (TDKN) is simply the
fraction of TKN that passes through a 0.45ym fil-
ter, separating out the particulate-N. Ammonium
(NH4) is another parameter commonly measured
and provides a breakdown of the TKN into inor-
ganic and organic N.

Total-P represents all organic and inorganic
forms of P, measurable using a standard chemical
analysis procedure (Izuno et al., 1991a). Other P
parameters commonly analyzed for are Ortho-Phos-
phate (PO,), Soluble Reactive Phosphorus (SRP),
and Total Dissolved Phosphorus (TDP). Particulate
P is calculated as the difference between TP and

The current SWIM Plan (SFWMD, 1990) focuses
on TP. Hence, the minimal analysis of a water
sample for south Florida should include TP. How-
ever, TP includes particulate forms of P, the
amount of which could be greatly influenced by
sampling procedures. Hence, it is advisable to also

analyze samples for TDP. Although Ortho-P and
SRP are the bio-available forms of P which would
have the most immediate effect on aquatic systems,
they are of limited value if monitoring is being con-
ducted for the purpose of complying with terms of
the SWIM Plan.

Other monitoring concerns
The SWIM Plan mandates a decrease in P load-
ing to the EAA agricultural drainage water receiv-
ing areas. In calculating the desired loading reduc-
tions, however, much attention was directed to P
concentrations. Loading is simply a flow-weighted
concentration multiplied by the volume of drainage
water pumped. Hence, loading reductions can be
achieved by reducing P concentrations and/or the
volume of drainage water pumped off a farm. How-
ever, water supply is also a major problem in south
Florida, placing limitations on the desirability of
retaining much water on farms. It is clear that a
farm level water quality monitoring program must
include measurements of the volume of drainage
water pumped off farms as well as the P concentra-
tion of that water.

Ideally, a monitoring program should enable a
grower to account for all water and P that enters
and leaves the confines of the farm. This informa-
tion will yield farm water and P budgets. However,
for SWIM Plan compliance purposes, only TP out-
puts to area canals will be necessary. To achieve
total farm water and P budgets, the following pa-
rameters will need to be obtained: 1) fertilization
amounts; 2) irrigation volumes and corresponding
P concentrations; 3) rainfall amounts and P concen-
trations; 4) drainage volumes and P concentrations;
and 5) evapotranspiration volumes. Irrigation and
rainfall concentrations are important since P con-
centrations in both of these waters, at times, exceed
P concentrations desired at the receiving areas
(Izuno et al., 1990a; LOTAC II, 1990; SFWMD,
1990). Additionally, even low P concentrations in
these waters can result in high loadings because of
the large volumes of water involved.

Finally, a monitoring program should also in-
clude farm water table levels in order to better de-
termine the effectiveness of the water management
programs for given farms or fields. Procedures for
accomplishing this will not be discussed in this
publication since they are covered in depth else-
where (Izuno et al., 1988; 1989a; 1989b; 1990).

Types of water samples
There are essentially three basic sampling pro-
cedures used in evaluating the effects of agricul-
tural production on water quality and the environ-
ment. Each sampling procedure has specific ad-
vantages and disadvantages, depending on the in-
tended use of the resulting data. The three sam-
pling protocols are: 1) point-in-time grab sam-
pling; 2) time sequenced sampling using
autosamplers; and 3) flow integrated proportional
sampling using integrating composite samplers.

Grab samples
Grab samples are water samples which are
manually collected from the pump sump, canal,
ditch, etc. and are representive only of one point in
time. The manual sample collection procedure re-
quires that an individual be present to physically
collect the sample. The sample collector may use
peristaltic pumps (very low flow positive displace-
ment pumps) or the sample may be simply dipped
or scooped out of the channel using a bottle. Data
inconsistencies can be introduced if the samples
are not collected at the same location each time.
To alleviate this problem, mark the sampling loca-
tion and depth or use a tube/strainer assembly (a
device used to restrict the uptake of particles that
would plug the suction hose or pump) that can be
permanently mounted at or near the pump station
in the pump sump, ditch, or canal. The hose at-
tached to the strainer can be run to a point above
the highest water surface level expected and
hooked to a pump when a sample is being col-

The most appropriate depth for sample collec-
tion is about 20 to 40 percent of the total channel
depth beneath the water surface. For systems that
do not use the fixed mounted strainer assembly, a
suction strainer can be attached to a length of hose
and lowered into the water body prior to start-up
of the sampling pump. In the case where a bottle
is used to dip a sample, the collector may simply
turn the bottle upside down and push it down to
the appropriate depth. No water will enter the
bottle since the above action traps the air inside
the bottle, inhibiting the air displacement with wa-
ter. At the appropriate depth, the bottle should be
tipped sideways or upright to allow water entry.
Once filled, the bottle may be capped under water
or raised to the surface sideways to reduce the po-
tential for contamination by water from undesired

depths. The obvious shortcoming of dipping
samples is that getting a sample from the proper
place within a canal can be an adventure and may
require an extremely long-armed individual with
great intestinal fortitude.

Theoretically, to delineate the P concentration
distribution over a pumping event, a series of grab
samples may be collected every hour or two during
the event. This can be accomplished by a very dedi-
cated individual. However, the procedure would be
time consuming and requires that a person be lo-
cated at the site throughout the drainage event.

In many water quality monitoring programs,
grab samples are collected on a preset time basis.



Z 10.0



4- +tHI

Hrl H Arllr


In other words, they are pulled daily, weekly,
monthly, etc., depending on the resolution desired.
However, such an approach is not appropriate for
our purposes. The water of interest in the EAA is
that which is pumped off the farm (drainage) and
that which enters a farm (irrigation). Periods when
neither irrigation nor drainage are occurring (i. e.
static water) are not of interest and samples should
not be gathered at those times. Including samples
from static water conditions in a data set can inap-
propriately bias average concentrations. An ex-
ample of monthly grab sample concentrations com-
pared with event average concentrations is shown
in Figure 1.

I I '


Figure 1. Monthly average total phosphorus concentrations from autosampled drainage events and monthly grab samples for veg-
etable field.


. . . ... . ...... ..... .... .. .....................

.. ....................................................?................. ................ ................

,, i IIIIi ,rii






Because nutrient concentrations vary greatly
during a pumping event (Figure 2), it is important
to be consistent in selecting when during the drain-
age event the grab samples are collected. Therefore,
samples should be collected in such a way that they
are representative of the entire drainage event.
Determining this sampling strategy will require the
intensive monitoring of several events to learn the
variability trends prior to reducing sample num-
bers greatly. Even with the establishment of the
nutrient concentration trends over time at the
pump station, there will still be a need to sample
several times during the pumping event to yield an
average that is more representative than a single
sample. One may be tempted to sample the drain-
age event at pump start-up, shut down, and mid-
way through the event. However, without intensive
sampling (defining the nutrient concentration time
series throughout the entire event) there is still no
guarantee that this "average" concentration will be
an accurate reflection of actual conditions. Note
that the other two sampling procedures reduce this

The primary advantage of grab samples is that
very little equipment is required for sample collec-
tion and flexibility in sampling location selection.
However, this method sacrifices data resolution due
to the small number of samples collected. The user
must also be aware that P concentrations change
during an event and that grab sampling will yield
no indication of whether the sample is representa-
tive of the event.

Autosamplers are instruments that are com-
posed of a timer, pump, sample distributor, and
sample bottles. They can be programmed to collect
volumes of water on desired time intervals, begin-
ning at a specific time. They place the water
sample in one of up to 24 bottles stored in the
autosampler base. Running from the autosampler
pump is a suction hose connected to a strainer
which is generally mounted in a specific position in
a ditch, canal, sump, etc. where water is being





0.35 -


0.25 -


0.15 -


v. IV
1000 14:00 18:00 22:00 02:00 06:00 10:00 14:00 18:00 22:00 02:00 06:00
Figure 2: Phosphorus concentration distribution over time during a drainage event at a sugarcane farm.


ll l l l lfI

sampled. Autosamplers are discussed in detail in
Taylor et al. (1991).

Autosamplers are designed to collect a series of
water samples throughout the duration of a drain-
age event without necessitating the presence of a
person. They can also be connected to pump con-
trols so that sampling initiation can be triggered by
pump operations. This type of sampling scheme is
by far the best for characterizing drainage water
nutrient concentrations. However, capital and
operating costs are relatively high.

As was shown in Figure 2, concentrations can
change during pumping. Collecting a series of
samples on relatively short, equal time intervals
during the drainage event ensures that a represen-
tative time averaged nutrient concentration re-
sults. Additionally, when using water sample
analyses to determine the effectiveness of a BMP
program, it is desirable to know when during the
drainage event phosphorus loading reductions oc-
curred as well as how much of a reduction was
achieved. Knowing the timing, duration, and
height of the nutrient concentration peaks under
BMP and pre-BMP conditions could yield important
differences that may have been hidden in a simple
average concentration for the event.

Flow integrating composite
Flow integrated composite samplers collect
water samples in amounts proportional to flow
rates during a pumping event. The water samples
are deposited in a single large container, yielding a
composite sample that accounts for flow changes
that occur during sampling.

A flow integrating sampler usually requires that
a water flow measuring device be installed in the
channel being monitored. The flow meter is elec-
tronically connected to the autosampler and a data
logger, and its output is used to control the sample
flow rate into the receiving bottle. These systems
are more complex with respect to installation and
management, and are expensive.

Flow integrated composite samples can also, at
times, be obtained when pumps are being used. A
good composite sample can be collected by simply
attaching a small tube to the discharge port of the
pump. The head on the discharge side of the pump
will be sufficient to produce a small stream of water

through the tube as long as the discharge pipe is
flowing full. The flow from the tube will be ap-
proximately flow proportional and can be deposited
in a large (5 to 10 gal) collection container. If the
flow fills the container too quickly, a smaller diam-
eter tube can be used or the sample stream can be
split on the edge of a funnel.

A single subsample can be extracted from the
large composite sample for analysis. It is probably
desirable to analyze more than one subsample be-
cause of the concentration variability between
subsamples which will occur when analyzing for
TP. The TP concentration of the subsample is flow
weighted, and hence, it can be directly multiplied
by the flow volume to calculate TP loads.

Sample collection locations
At the very least, a water sample collection sta-
tion should be located near the main pump station
serving the farm. Ideally, samples should be taken
directly from the pump discharge pipe, before it has
a chance to mix with canal water outside the farm
boundaries. In the case where the discharge pipe is
above the canal or discharge sump water level,
sampling discharge water is a fairly easy task to
accomplish. In the case of a submerged outlet pipe,
water should be taken from within the discharge
pipe, necessitating access into the pipe. Simply
placing the autosampler suction strainer in the dis-
charge pit may yield highly variable results since
the bottom sediment will be stirred up by the dis-
charge waters.

An autosampler could be placed at the pump in-
let in the sump pit. The operator will have to con-
tend with floating trash and rapidly fluctuating wa-
ter levels. Additionally, much sediment will be
stirred up by the vortex created as water is sucked
up into the pump. On the other hand, the pump
inlet sump pit should provide a well mixed water
sample. The operator may also have to install an
additional sampling station to account for irrigation
water unless it feeds into the drainage inlet sump.

A sampling station can be installed upstream of
the pump within the confines of the farm (Figure
3). The station should be at a point where no water
can enter or leave the farm in surface ditches with-
out passing the sampler. Mounting the sampler a
sufficient distance upstream of the pump station
will minimize the rapid and frequent changes in
water level, enabling the collection of samples from

a more identifiable point in the flowing water. This
will also eliminate the need to tap into the pump
discharge pipe. The obvious shortcoming of this
placement is that some of the sediment passing
through the pump will not be picked up in the
Each of the three location arrangements has ad-
vantages and disadvantages. The important thing
to remember is to be consistent and straightfor-
ward in reporting results. In other words, once a
monitoring strategy is set up, it should be clearly
reported and remain in place for the duration of
sampling activities.

Strainers should be located at the point in the
canal cross section representative of the maximum
water velocity in the channel. The flow velocity
profile approximates a paraboloid, with zero flow
velocity occurring at the channel sides and bottom
and maximum velocity in the center of the canal
slightly below the water surface. However, strain-
ers should be placed deep enough below the water
surface to avoid interference from floating plants
and debris. They should also be maintained far
enough above the channel bottom to prevent sedi-
ment from being sucked up into the sample. Ideal
placement is, therefore, in the middle of the chan-
nel, 20 to 40 percent of the depth below the surface.


mLIT -


i I I I I
! i i i

~I .
I 1 i i !
I u,.c I

I i I

Note that the channel depth lowers during pump-
ing, sometimes dramatically. Hence, locating the
strainer as suggested will avoid detrimental water
surface and channel bottom conditions during the
majority of events.

Nutrient budgets
Water quality monitoring can proceed along one
of two lines: 1) nutrient concentrations or 2) nutri-
ent loading. Both are important in determining the
effects of agricultural production on the environ-
ment as well as determining the effects of BMPs on
phosphorus reductions in drainage water. Calcu-
lating nutrient loading is, however, dependent on
having nutrient concentration data. The end prod-
uct of a nutrient loading determination strategy is
the development of a nutrient budget.

Nutrient budgets require that flow volumes into
and out of the farm be measured during irrigation
and drainage, respectively. These flow volumes,
when multiplied by flow-weighted nutrient concen-
trations, will yield the mass of nutrients entering
or leaving a farm in water.

Water serves as the primary pathway of nutri-
ents into or out of a farm, other than fertilization

i ii

i l i
! i
_. : i i


i i I I

i i i i

I i i i i
l i I i i


Figure 3: A suggested water sampling scheme for a hypothetical farm.

Farm Nutrient Budget
Change in Storage = Precipitation + Irrigation + Seepage in + Overland Flow in -
Drainage Seepage out Overland Flow out
+ Fertilization Harvest +/- Intangibles






< 0==





Figure 4: Farm nutrient budget illustrating the directions of nutrient movement and its dependence on water flows.

and crop removal (Figure 4). All flows and nutrient
concentrations can be measured, except for under-
ground seepage into the rock profile and lateral
seepage to other farms or area canals. In cases,
seepage can be a major phenomenon affecting the
ability to account for water volumes and, therefore,
nutrient budgets. Seepage volumes are best esti-
mated as the water that cannot be accounted for
after measuring all other inflows and outflows in-
cluding evapotranspiration which can be estimated
from pan evaporation.

The nutrient concentration to attribute to seep-
age is dependent on many complex physical and
chemical factors. Hence, rough estimates are
made, or generally, seepage is simply ignored. In
most cases in the EAA, the assumption that seep-
age is a minor element in the water and nutrient
budgets is plausible. In other cases, it can be com-
pletely erroneous. Generally, the larger the farm
area in relation to its length of boundaries along
major off-farm canals, the better the assumption.

The intangibles include other sources or sinks of
nutrients that could have major effects on the farm
budget. A good example is the case of rice fields

where birds and fish enter the fields and greatly
alter the nutrient pathways by eating and deposit-
ing organic matter.

Selecting events to sample

Drainage monitoring
Selecting drainage events to sample poses ques-
tions that are not easily answered. One is not able
to predict the number and intensity of storms and
how much water will be pumped during each
month of the year. Collecting and analyzing
samples is expensive (Taylor et al., 1991), preclud-
ing the extensive monitoring of every drainage
event of the year. The logistics of handling thou-
sands of samples can be time consuming and con-
fusing. Sampling should, therefore, be kept to a
manageable level.

There are ways to minimize the sample collec-
tion and analysis load. For example, samples can
be collected on preset time intervals throughout the
event. A few samples can then be selected for ini-
tial analysis while the others are preserved and
stored for future analysis if it is determined that



%nQM1 %c~4h.

further resolution is needed after seeing the results
of the first samples analyzed. This reduces the
amount of laboratory work and expense while en-
suring that the nutrient concentration time series
is available if desired.

Alternatively, if the operator is not concerned
with the behavior of nutrient concentrations over
time, sample load can be reduced by using a
compositing autosampler. Individually collected
samples can be composite in a flow-weighted man-
ner by simply using flow records collected during
the event and combining flow proportional amounts
of sample from each bottle.

Finally, one can eliminate the collection of time
sequenced samples by using a flow integrating com-
posite autosampler. In this case, a single flow pro-
portioned sample is collected during a drainage
event using an appropriately configured

In addition to sampling at the main pump sta-
tions, a grower may wish to sample water at points
in the interior of a farm. Likely locations would be
at discharge and inlet points to blocks of land dedi-
cated to a particular crop. For example, ditches
serving vegetable seedbeds or production areas, or
rice fields, within a sugarcane farm would be ideal
sampling sites. A schematic of a typical water
quality monitoring arrangement for a hypothetical
farm is shown in Figure 3. These water samples
from sites internal to the farm would yield valuable
information for establishing background concentra-
tions and for making BMP or other crop manage-
ment comparisons. They would, however, have
little use if the only desire was to monitor compos-
ite farm discharges and inflows.

Drainage water nutrient concentrations vary
through the year and with rainfall intensity (Izuno
et al., 1990a) as was shown in Figure 1. Although
there are definite rainy and dry seasons, drainage
pumping can occur throughout the year. Hence, it
is important to monitor nutrient concentrations
over the entire year, not just during a few isolated
rainstorms or drainage events randomly scattered
or tightly grouped in time. In other words, collect-
ing hundreds of samples during a couple of drain-
age events following major rainfalls during the
peak of the rainy season, while ignoring drainage
events during the dry season, will yield an errone-
ous picture of the nutrient concentrations in EAA
drainage water. In fact, the error will generally be
on the high side. Likewise, sampling during short

drainage events during the dry season will not ad-
equately characterize the farm nutrient discharge
for the year. It is obvious that a comprehensive,
well thought out monitoring program is mandatory
to obtain accurate information.

Irrigation monitoring
While public attention is focused on the quality
of drainage water, monitoring the quality and
quantity of irrigation water is equally important.
Area canals contain nutrients in concentrations
that vary according to hydrologic and hydraulic
conditions. A farm nutrient budget must account
for how much nutrient enters a farm in irrigation
to avoid misconceptions that all nutrients in farm
drainage water originate on the farm. While fertil-
izer and mineralization are looked upon as obvious
nutrient sources, irrigation water can contribute
significantly to the nutrient input to the farm, espe-
cially since large amounts of water may be let into
the farm during a year. In some cases, irrigation
water is pumped or gravity fed into a farm and
then pumped right back out when an unexpected
rainfall occurs. Likewise, a farm may be drained in
anticipation of a major rainfall and then irrigated
when the rainfall didn't materialize or was of lesser
intensity than expected. To obtain an accurate pic-
ture of a farm's net P balance and related nutrient
loading problems, irrigation water should be moni-

In many instances, the same installation used
for collecting drainage water samples can be used
for sampling irrigation events. If irrigation water
is let into the farm at a point at or near the pump
station, an autosampler located upstream of the
pump station will suffice. A collection installation
in the drainage pipe cannot be used for irrigation
water sampling unless backsiphoning or reverse
pumping is used for irrigation. Depending on the
hydraulic arrangement at the inlet of the pump sta-
tion with respect to irrigation and drainage facili-
ties, a sampling station in the drainage inlet sump
may or may not suffice.

Soil solution sampling
Sampling the soil solution, or water stored in the
soil profile for nutrient analysis serves a unique
purpose. Analysis of these samples can indicate
the amount of phosphorus in the soil solution and
whether it is increasing or decreasing under differ-
ent BMPs imposed. This information is useful dur-
ing BMP evaluation.

Soil solution samples are extracted from the soil
through instruments called solution samplers or
suction lysimeters. These instruments and sam-
pling procedures are described in detail in Taylor et
al. (1991).

Rainwater samples
Rainfall contributes significant amounts of nutri-
ents and water to a farm over the course of a year
and therefore, must be monitored to prepare farm
water and nutrient budgets. It is important that
rainfall collectors for nutrient analyses be properly
prepared (Taylor et al., 1991). Rainfall samples
must not be taken from the raingage for analysis.
Generally, the rainwater sample will contain dry
fallout (particulate matter) that occurs coincident
with the rainfall. Hence, the sample is often re-
ferred to as bulk precipitation (wet and dry fallout)
that occurred during a rainfall. Bulk precipitation
is a better representation of the total atmospheric P
input than just rainfall itself. Rainfall should be
measured at one or more sites around a farm, de-
pending on the farm size. However, nutrient load-
ing attributed to bulk precipitation can be ap-
proached on a regional scale due to the low vari-
ability between precipitation nutrient concentra-
tions in similar areas.

Recommended water quality
monitoring strategy
A detailed water quality monitoring strategy
may be prohibitively expensive with respect to
time, instrumentation, and sample analyses. The
challenge is, therefore, to develop a strategy that is
a compromise between the ideal and the
layperson's tendency to simply dip a sample occa-
sionally. It is important to plan a water sampling
program carefully, document it rigorously, and con-
duct it conscientiously. It should be remembered
that the monitoring program and resulting data
may be important for on-farm management, scien-
tific, and legal reasons.

A reasonable program could begin with locating
sampling stations upstream of the main farm pump
stations that access area canals. Autosamplers
could be used, with suction strainers placed in the
middles of the farm canals, 20 to 40 percent of the
channel water depth, below the water surface. All
farm inflows and outflows should pass by a moni-
toring station. Water level recorders should be in-

stalled at the water quality sampling sites. These
recorders can be used for measuring flow volumes
or simply for automatic recording of irrigation and
drainage event times. Pump stations should be
calibrated such that flow volumes can be calcu-
lated. Ideally, inflow structures should also be cali-
brated to determine the volume of water entering
the farm.

In cases where detailed time sequenced water
samples are not desired or are unnecessary, flow
integrating autosamplers would probably be the
most affordable and manageable option.

An intensive water sampling program should be-
gin with a program to assess when peak concentra-
tions occur during drainage and irrigation events.
Samplers should be set to collect water every one to
two hours, depending on the expected duration of
the event. If the event lasts longer than antici-
pated and the supply of bottles is used up, simply
change the autosampler base with a fresh set of
bottles. Select a set of 24 samples, equally spaced
in time, for analysis. Other samples may be stored
or discarded. After following this procedure a few
times, the grower will be able to determine the few-
est number of samples needed to characterize the
water quality at the station during events by exam-
ining graphs of water quality versus time. Inten-
sive monitoring should be done for several events
during different times of the year in order to char-
acterize small and large rainfall events during the
rainy and dry seasons.

When interpreting the water quality data, cau-
tion is strongly advised. The correct interpretation
of water quality monitoring data can be complex,
requiring an expert in statistical analyses to test
for significant differences and trends. The proper
statistical interpretation of data is important in the
assessment of the efficacy of a BMP implementa-
tion program.

Irrigation, drainage, and rainfall samples should
be analyzed for total phosphorus (TP) and total dis-
solved phosphorus (TDP) at the very least.

There are many factors to consider when install-
ing water quality monitoring stations on farms.
While there may not be any absolutely "right" way,
the grower should consider his farm layout, crop-
ping pattern, and desired information prior to

implementing an appropriate program. Consis-
tency in sampling location and technique is ex-
tremely important. Sampling strategies and data
should be carefully and precisely documented. Flow
volumes should also be monitored to attain infor-
mation necessary for developing farm nutrient bud-
gets. Using autosamplers during drainage events
that occur throughout the year will yield the most
detailed assessment of nutrient concentrations and
loadings. However, flow integrated sampling would
be the most cost effective means for determining
net loadings and flow averaged concentrations. Fi-
nally, a word of caution. As is often the case, a
little information from a poorly designed monitor-
ing program can be more damaging than having no
information at all. Growers are therefore advised
to seek professional help in developing monitoring
program designs and sampling strategies.

Florida Department of Administration. 1976.
Final report on the special project to prevent eu-
trophication of Lake Okeechobee. 341 pp. Novem-
Izuno, F. T. and A. B. Bottcher, eds. 1991. The
effects of on-farm agricultural practices in the or-
ganic soils of the EAA on phosphorus and nitrogen
transport. Final Report submitted to the South
Florida Water Management District, West Palm
Beach. March.
Izuno, F. T., A. B. Bottcher, F. J. Coale, C. A.
Sanchez, and D. B. Jones. 1990a. Agricultural best
management practices (BMPs) for phosphorus load-
ing reduction in the Everglades Agricultural Area
(EAA). Preliminary Final Report submitted to the
South Florida Water Management District. 42 pp.
Izuno, F. T., G. A. Clark, D. Z. Haman, A. G.
Smajstrla, and D. J. Pitts. 1989a. Manual moni-
toring of farm water tables. IFAS Extension Circu-
lar 731. Florida Cooperative Extension Service, In-
stitute of Food and Agricultural Sciences, Univer-
sity of Florida, Gainesville. 19 pp. June.
Izuno, F. T., G. A. Clark, F. S. Zazueta, and D. Z.
Haman. 1990b. Electronic methods for agricultural

water level monitoring. IFAS Extension Bulletin
264. Florida Cooperative Extension Service, Insti-
tute of Food and Agricultural Sciences, University
of Florida, Gainesville. 22 pp. May.
Izuno, F. T., W. A. Davis, and A. B. Bottcher.
1991a. Terms related to agricultural nutrient load-
ing in south Florida. IFAS Extension Circular 945.
Florida Cooperative Extension Service, Institute of
Food and Agricultural Sciences, University of
Florida, Gainesville.
Izuno, F. T., D. Z. Haman, and G. A. Clark.
1988. Water table monitoring. IFAS Extension
Bulletin 251. Florida Cooperative Extension Ser-
vice, Institute of Food and Agricultural Sciences,
University of Florida, Gainesville. 23 pp. August.
Izuno, F. T., D. Z. Haman, and G. A. Clark.
1989b. Strip chart recorders: Applications in on-
farm water management. IFAS Extension Bulletin
253. Florida Cooperative Extension Service, Insti-
tute of Food and Agricultural Sciences, University
of Florida, Gainesville. 23 pp. June.
Izuno, F. T., C. A. Sanchez, F. J. Coale, A. B.
Bottcher, and D. B. Jones. 1991b. Phosphorus con-
centrations in drainage water in the Everglades
Agricultural Area. Journal of Environmental Qual-
ity 20(3):608-619.
LOTAC I. 1986. Final Report. Submitted to the
Governor of the State of Florida, the Secretary of
the Department of Environmental Regulation, and
the Governing Board of the South Florida Water
Management District. Tallahassee.
LOTAC II. 1990. Final Report. Submitted to
the Governor of the State of Florida, the Secretary
of the Department of Environmental Regulation,
and the Governing Board of the South Florida Wa-
ter Management District. Tallahassee.
SFWMD. 1990. Surface water improvement and
management plan for the Everglades, Final Draft.
South Florida Water Management District, West
Palm Beach. September.
Taylor, L. A., F. T. Izuno, and A. B. Bottcher.
1991. Water quality sampling and analysis instru-
ments. IFAS Extension Circular 1040. Florida Co-
operative Extension Service, Institute of Food and
Agricultural Sciences, University of Florida,

Director, in cooperation with the United States Department of Agriculture, publishes this information to further the purpose of the May 8 and June
30,1914 Acts of Congress; and is authorized to provide research, educational information and other services only to individuals and institutions that
function without regard to race, color, sex, age, handicap or national origin. Single copies of extension publications (excluding 4-H and youth
publications) are available free to Florida residents from county extension offices. Information on bulk rates or copies for out-of-state purchasers
s availablefrom C.M. Hinton, Publications Distribution Center, IFAS Building 664, University of Florida, Gainesville, Florida 32611. Before publicizing
this publication, editors should contact this address to determine availability. Printed 1/92.

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs